U.S. patent number 10,527,365 [Application Number 16/030,097] was granted by the patent office on 2020-01-07 for disconnect assembly for active cooling of packaged electronics.
This patent grant is currently assigned to International Business Machines Corporation. The grantee listed for this patent is International Business Machines Corporation. Invention is credited to Thomas Brunschwiler, Ingmar G. Meijer, Stephan Paredes, Gerd Schlottig.
![](/patent/grant/10527365/US10527365-20200107-D00000.png)
![](/patent/grant/10527365/US10527365-20200107-D00001.png)
![](/patent/grant/10527365/US10527365-20200107-D00002.png)
![](/patent/grant/10527365/US10527365-20200107-D00003.png)
![](/patent/grant/10527365/US10527365-20200107-D00004.png)
![](/patent/grant/10527365/US10527365-20200107-D00005.png)
![](/patent/grant/10527365/US10527365-20200107-D00006.png)
![](/patent/grant/10527365/US10527365-20200107-D00007.png)
![](/patent/grant/10527365/US10527365-20200107-D00008.png)
![](/patent/grant/10527365/US10527365-20200107-D00009.png)
United States Patent |
10,527,365 |
Schlottig , et al. |
January 7, 2020 |
Disconnect assembly for active cooling of packaged electronics
Abstract
A disconnect assembly includes a solid frame comprising a slit
and a first liquid coolant circuit leading to a frame outlet
defined in an inner wall of the slit. The assembly further includes
an insert element, insertable in the slit so as to reach a sealing
position. The latter defines a shut state, in which the insert
element seals the frame outlet. The assembly includes a cold plate,
comprising a second liquid coolant circuit with a duct open on a
side of the cold plate. The cold plate can be inserted in the slit,
so as to push the insert element, for the latter to leave the
sealing position and the cold plate to reach a fluid communication
position. This position defines an open state, in which the duct is
vis-a-vis the frame outlet, to enable fluid communication between
the first liquid coolant circuit and the second liquid coolant
circuit.
Inventors: |
Schlottig; Gerd (Uitikon
Waldegg, CH), Paredes; Stephan (Zurich,
CH), Meijer; Ingmar G. (Zurich, CH),
Brunschwiler; Thomas (Thalwil, CH) |
Applicant: |
Name |
City |
State |
Country |
Type |
International Business Machines Corporation |
Armonk |
NY |
US |
|
|
Assignee: |
International Business Machines
Corporation (Armonk, NY)
|
Family
ID: |
69058545 |
Appl.
No.: |
16/030,097 |
Filed: |
July 9, 2018 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H05K
7/20309 (20130101); F28F 9/0258 (20130101); F28F
9/0251 (20130101); H05K 7/20327 (20130101) |
Current International
Class: |
F28D
1/00 (20060101); H05K 7/20 (20060101); F28F
9/02 (20060101) |
References Cited
[Referenced By]
U.S. Patent Documents
Other References
Goldman, R., et al., "Designing a Liquid Cooling Loop for
High-Performance Systems", LYTRON Total Thermal Solutions,
http://www.lytron.com/Tools-and-Technical-Reference/Application-Notes/Des-
igning-a-Liquid-Cooling-Loop-for-High-Performance-Systems, Accessed
on Jul. 6, 2018, 3 pages. cited by applicant .
Kelkar, K.M., et al., "Analysis and Design of Liquid-Cooling
Systems Using Flow Network Modeling (FNM)",
http://inresllc.com/assets/files/macroflow/MF03-Design_of_Liquid_Cooling_-
Systems.pdf, Proceedings of IPACK03 International Electronic
Packaging Technical Conference and Exhibition, Jul. 6-11, 2003, 6
pages. cited by applicant.
|
Primary Examiner: Gandhi; Jayprakash N
Assistant Examiner: Matey; Michael A
Attorney, Agent or Firm: Scully, Scott, Murphy &
Presser, P.C. Morris, Esq.; Daniel P.
Claims
What is claimed is:
1. A disconnect assembly for active cooling of packaged
electronics, the assembly comprising: a solid frame comprising a
slit and a first liquid coolant circuit leading to a frame outlet
defined in an inner wall of the slit; an insert element, insertable
in the slit so as to reach a sealing position defining a shut
state, in which the insert element seals the frame outlet; and a
cold plate, comprising a second liquid coolant circuit with a duct
open on a side of the cold plate, wherein the cold plate is
insertable in the slit so as to push the insert element for the
latter to leave the sealing position and the cold plate to reach a
fluid communication position defining an open state, in which the
duct is vis-a-vis the frame outlet, to thereby enable fluid
communication between the first liquid coolant circuit and the
second liquid coolant circuit.
2. The disconnect assembly according to claim 1, further comprising
a gasket arrangement extending on both the insert element and the
cold plate, the gasket arrangement designed to prevent liquid
leakage from the insert element and the cold plate in each of said
shut state and said open state.
3. The disconnect assembly according to claim 2, wherein the gasket
arrangement is further designed to prevent liquid leakage from the
insert element and the cold plate during a transition from said
shut state to said open state.
4. The disconnect assembly according to claim 3, wherein the gasket
arrangement comprises a toric joint, located on the cold plate,
wherein the toric joint surrounds an aperture of the duct open on
said side of the cold plate, so as to surround the frame outlet if
the cold plate is in said fluid communication position and prevent
liquid leakage from the frame outlet in the open state.
5. The disconnect assembly according to claim 4, wherein said toric
joint is a second toric joint, the gasket arrangement comprising a
first toric joint located on the insert element, so as to surround
the frame outlet if the insert element is at said sealing position
and prevent liquid leakage from the frame outlet if the assembly is
in said shut state.
6. The disconnect assembly according to claim 5, wherein the gasket
arrangement further comprises: an outer toric joint divided in two
halves, comprising: a first half located on the insert element, on
a same side as the first toric joint, the latter partly surrounded
by the first half of the outer toric joint on that same side; and a
second half located on the cold plate on a same side as the second
toric joint, the latter partly surrounded by the second half on
that same side, whereby the two halves meet and form the outer
toric joint if the cold plate contacts the insert element in the
slit, so as for the first toric joint and the second toric joint to
be located within an area bounded by the outer toric joint.
7. The disconnect assembly according to claim 6, wherein the
assembly further comprises fastening means configured to help
maintaining said two halves forming the outer toric joint, in
operation.
8. The disconnect assembly according to claim 6, wherein the cold
plate further comprises: a liquid reservoir; and one or more
apertures open on said side of the cold plate, outside the second
toric joint, so as for the one or more apertures to be surrounded
by the outer toric joint, and wherein said liquid reservoir is in
fluid communication with said one or more apertures, via a third
liquid coolant circuit that is independent from said second liquid
circuit, so as to evacuate liquid coming from the frame outlet
toward the reservoir during a transition from said shut state to
said open state, in operation of the assembly.
9. The disconnect assembly according to claim 8, wherein said
reservoir is designed so as to allow evaporation of liquid
accumulated therein via an edge surface of the cold plate.
10. The disconnect assembly according to claim 9, wherein the cold
plate further comprises a liquid evaporation medium extending along
an edge of the cold plate, said evaporation medium interfacing a
volume defined in said reservoir with an external environment
medium to the cold plate.
11. The disconnect assembly according to claim 6, wherein a face of
the cold plate, on which said second toric joint is arranged, is
shifted along a normal of that face with respect to a face of the
insert element on which said first toric joint is arranged.
12. The disconnect assembly according to claim 6, wherein each of
said two halves of the outer toric joint, the first toric joint and
the second toric joint is partly embedded under a respective
anchorage surface of the insert element or the cold plate.
13. The disconnect assembly according to claim 1, wherein the cold
plate further comprises liquid cooling structures arranged in a
volume defined within the cold plate, at an end of the second
liquid coolant circuit.
14. The disconnect assembly according to claim 1, wherein each of
the solid frame and the cold plate are symmetrically designed and
the assembly further comprises a second insert element, whereby the
first liquid coolant circuit of the solid frame comprises two
disconnected subsections and the cold plate is insertable in a pair
of opposite slits of the solid frame, so as to push insert elements
therein and connect said subsections via the second liquid coolant
circuit.
15. The disconnect assembly according to claim 14, wherein the
first liquid coolant circuit and the second liquid coolant circuit
form parts of a liquid cooling circuit pathing through the solid
frame and the insert element, which liquid cooling system is, in
said open state, a closed liquid circuit.
16. The disconnect assembly according to claim 1, wherein the solid
frame comprises: n slits and n corresponding first liquid coolant
circuits leading to respective frame outlets, the latter defined,
each, in an inner wall of a respective one of the n slits, n>1;
n insert elements, wherein each of the insert elements is
insertable in a respective one of the n slits so as to reach a
sealing position defining a shut state, in which said each of the n
insert elements seals a respective one of the frame outlets; and n
cold plates with n second liquid coolant circuits, each of the cold
plates comprising a respective duct open on a side thereof, the
duct leading to a respective one of the n second liquid coolant
circuits, wherein each of the cold plates is insertable in a
respective one of the slits so as to push a respective one of the
insert elements, for the latter to leave its sealing position and
said each of the cold plates to reach a fluid communication
position that defines an open state, in which a respective duct is
vis-a-vis one of the frame outlets to enable fluid communication
between one of the n first liquid coolant circuits and one of the
second liquid coolant circuit.
17. A hardware system comprising the disconnect assembly according
to claim 1, and packaged electronics mounted on the cold plate, to
thereby cool down the packaged electronics, in operation of the
system.
18. A method of operation of an assembly for active cooling of
packaged electronics, wherein the assembly comprises a solid frame
comprising a slit and a first liquid coolant circuit leading to a
frame outlet defined in an inner wall of the slit, the method
comprising: inserting an insert element in the slit so as for it to
reach a sealing position defining a shut state, in which the insert
element seals the frame outlet; and inserting a cold plate in that
same slit, the cold plate comprising a second liquid coolant
circuit and a duct open on a side of the cold plate, the duct
leading to the second liquid coolant circuit, so as to push the
insert element for the latter to leave the sealing position and the
cold plate to reach a fluid communication position defining an open
state, in which the duct is vis-a-vis the frame outlet, to enable
fluid communication between the first liquid coolant circuit and
the second liquid coolant circuit.
Description
BACKGROUND
The present disclosure relates in general to the field of hardware
cooling and in particular to active cooling solutions for packaged
electronics. Embodiments of the invention are directed to
assemblies of parts that can be inserted in slits of a solid frame
housing a liquid cooling circuit, in order to open or shut sections
of the cooling circuit.
Various approaches to cool hardware systems such as packaged
electronics have been proposed. For instance, several active
liquid-cooling solutions are known. In general, an active cooling
system makes it difficult, if not impossible, to hot-plug elements
of the system when the active liquid-cooling is on. Quick
disconnect solutions have been proposed for hot plugging. However,
such solutions are relatively complex and therefore expensive. In
all cases, leakage of coolant in liquid-cooled systems is a risk,
which can damage the hardware, especially while hot-plugging
elements on which the packaged electronics are mounted.
SUMMARY
According to a first aspect, the present invention is embodied as a
disconnect assembly for active cooling of packaged electronics. The
assembly includes a solid frame that comprises a slit and a first
liquid coolant circuit leading to a frame outlet defined in an
inner wall of the slit. The assembly further includes an insert
element, which is insertable in the slit so as to reach a sealing
position. The latter defines a shut state, in which the insert
element seals the frame outlet. Finally, the assembly includes a
cold plate, which comprises a second liquid coolant circuit with a
duct open on a side of the cold plate. The cold plate can be
inserted in the slit, so as to push the insert element, for the
latter to leave its sealing position and the cold plate to reach a
fluid communication position. The latter position defines an open
state, in which the duct is vis-a-vis the frame outlet, to enable
fluid communication between the first liquid coolant circuit and
the second liquid coolant circuit.
The invention may for example be embodied as a hardware system
comprising the above disconnect assembly, as well as packaged
electronics mounted on the cold plate, to thereby cool down
electronics in operation of the system.
According to another aspect, the invention is embodied as a method
of operation of an assembly for active cooling of packaged
electronics. The assembly includes a solid frame, a cold plate and
an insert element, as described above. According to this method,
the insert element is inserted in the slit so as for it to reach a
sealing position that defines a shut state, in which the insert
element seals the frame outlet. Next, the cold plate is inserted in
that same slit, so as to push the insert element, so as for the
latter to leave the sealing position and the cold plate to reach a
fluid communication position (open state), to enable fluid
communication between the first liquid coolant circuit and the
second liquid coolant circuit.
Devices, hardware systems and methods embodying the present
invention will now be described, by way of non-limiting examples,
and in reference to the accompanying drawings.
BRIEF DESCRIPTION OF SEVERAL VIEWS OF THE DRAWINGS
The accompanying figures, where like reference numerals refer to
identical or functionally similar elements throughout the separate
views, and which together with the detailed description below are
incorporated in and form part of the present specification, serve
to further illustrate various embodiments and to explain various
principles and advantages all in accordance with the present
disclosure, in which:
FIG. 1 is a 3D view of an assembly for active cooling of packaged
electronics device, according to embodiments;
FIGS. 2A-2D show various views of the cold plate of the assembly
depicted in FIG. 1 according to embodiments, wherein the affixed
packaged electronics are not depicted, for clarity;
FIG. 3 shows details of the cold plate in contact with an insert
element such as depicted in FIG. 1, as involved in embodiments;
FIGS. 4A-4E are 2D, orthographic projection views of elements of
the assembly of FIG. 1, illustrating the operation of such
elements, according to embodiments; and
FIG. 5 is a 3D view of an assembly having a symmetric design,
enabling a closed-loop cooling circuit, as in embodiments. Again,
the affixed packaged electronics are not depicted, for clarity.
The accompanying drawings show simplified representations of
devices or parts thereof, as involved in embodiments. Details are
sometimes omitted, for clarity. For example, the affixed packaged
electronics are not depicted in the drawings shown in FIGS. 2-5,
contrary to FIG. 1. Also, technical features depicted in the
drawings are not necessarily to scale. Similar or functionally
similar elements in the figures have been allocated the same
numeral references, unless otherwise indicated.
DETAILED DESCRIPTION OF EMBODIMENTS OF THE INVENTION
In reference to FIGS. 1 and 4, an aspect of the invention is first
described, which concerns a disconnect (or connect/disconnect)
assembly 1 for active cooling of packaged electronics 50.
Basically, this assembly 1 includes a solid frame 10 and at least
one set of movable elements, which set includes a cold plate 30 and
at least one insert element 20, to switch the cooling circuit. The
assembly 1, for example, has an essentially symmetric design with
respect to plane (x, z), as illustrated in FIG. 5, though this is
not strictly necessary. This aspect is discussed later in
detail.
The solid frame 10 comprises at least one slit 11, although several
slits will typically be involved, to allow dense package
arrangements. In the frame 10 is defined a first liquid coolant
circuit 12-14 (or circuit portion) for each slit 11, which circuit
leads to a frame outlet 14. The outlet 14 is defined in the inner
wall 11-z of the slit 11 meant to receive movable elements 20, 30.
Note, in the accompanying drawings, the inner walls of any slit 11
are identified according to indices corresponding to the normal
vectors to such walls. Thus, the walls 11.sub.z and 11.sub.-z
extend opposite to each other and perpendicularly to axis z.
Similarly, the wall 11.sub.x extends perpendicular to axis x, and
the wall 11.sub.-y extends adjacently between the walls 11.sub.z
and 11.sub.-z, etc.
Additional liquid circuit portions may otherwise be defined in the
solid frame 10, which may typically comprise ducts 13 opening
towards respective slits 11, so as to enable a coolant liquid to
pass into corresponding circuit portions 12-14, as seen in FIG. 1.
Each duct 13 may for instance communicate with a lateral channel
12, defined in the frame 10. Note, only one duct 13 and only one
outlet 14 are visible in FIG. 1, for reasons of concision.
The insert element 20 is designed so as to be inserted in the slit
11, e.g., by pushing it into the slit. The insert element 20 can
thus reach a sealing position in the slit 11. This sealing position
defines a shut state of the corresponding liquid cooling circuit
portion. In the sealing position, the insert element 20 seals the
frame outlet 14 and therefore shut this circuit portion.
A second liquid coolant circuit 31-33 (or circuit portion) is
defined in the cold plate 30. The circuit 31-33 notably includes a
duct 31, and an extension portion 32. The duct 31 is open on a side
of the cold plate 30, so as to enable liquid to enter the duct 31
from that side. The cold plate 30 too is designed so as to be
inserted in the slit 11. This way, the plate 30 may be brought in
contact with and push the insert element 20 (in the direction
opposite to that of axis x, see FIGS. 4A, 4B). As a result, the
insert 20 will leave its sealing position, while the cold plate 30
may reach a fluid communication position, in which the duct 31 is
(at least partly) vis-a-vis the frame outlet 14. This, in turn,
enables fluid communication between the first circuit portion 12-14
and the second circuit portion 31-33.
Note, a fluid communication position of a cold plate 30 defines an
open state of the liquid cooling circuit. Now, as the insert 20 and
plate 30 can be continuously moved along axis x in the slit 11, the
aperture of the duct 31 may only partly overlap (in projection)
with the aperture of the frame outlet 14 and thus only partly open
the circuit. When the axis of the duct 31 coincide with the axis of
the frame outlet 14, the overlap is maximal and the circuit is
fully open. There, the circuit can be said to be in a fully open
state. Still, one understands that there can in fact be several
"fluid communication positions" of the cold plate 30 along x. I.e.,
the assembly may be designed in such a manner that there is a
given, a finite interval of positions of the cold plate 30 for
which fluid communication is enabled, yet in an extent that depends
on the actual position of the plate 30.
Similarly, the sealing position may not be unique; there may be a
finite interval of positions of the insert 20 that all result in
sealing the circuit. However, there, the circuit is fully shut (or
closed) for all such positions, as we shall see. Therefore, the
overall circuit is either shut (by the insert 20), partly open
(thanks to the cold plate's duct 31 being at least partly vis-a-vis
the outlet 14) or fully open (when the axes of the duct 31 and
outlet 14 coincide).
The present solutions allow quick connect/disconnect of cold plates
30 in an active cooling system for packaged electronics 50. As seen
in FIG. 1, a packaged chip 50 may for example be mounted on one
side of the plate 30. In variants (not shown), two packaged
electronic components may be mounted on opposite sides of the plate
30. As further seen in FIG. 1, the solid frame 10 will likely
comprise several slits 11, all designed for receiving respective
inserts 20 and plates 30. The sets of slits 11, inserts 20 and
plates 30 may all be similarly designed (this is not a strict
requirement), so as to open or shut a liquid cooling circuit or a
subsection thereof.
FIGS. 1-4 shows only one pair of insert element 20 and cold plate
30, for the sake of depiction. In the following, particular
embodiments are described in reference to this pair of elements 20,
30, it being understood that the principles discussed herein can be
extended to multiple sets of elements 11, 20, 30, as well as
symmetric designs such as depicted in FIG. 5.
Referring now to FIGS. 1-3, the disconnect assembly 1, for example,
includes a gasket arrangement 40, which extends on both the insert
element 20 and the cold plate 30. This gasket arrangement 40 is
generally designed to prevent liquid leakage from the insert
element 20 and the cold plate 30, in each of the sealed and open
states of the corresponding liquid circuit, or circuit subsections.
As seen in FIG. 1, the gasket arrangement 40 extends on a same side
of the insert element 20 and the cold plate 30, when such elements
20, 30 are oriented so as to be inserted in the slit 11, to alter
the state of a corresponding cooling circuit.
A sophisticated gasket arrangement 40 is, for example,
contemplated, which allows each of the insert element 20 and the
cold plate 30 to hold stable in a slit 11 without exerting force
thereon (gravity will not appreciably impact the system's states
here), while allowing such elements 20, 30 to be pushed further
down in the slit, as necessary to switch from one state to the
other. Interestingly, and as present Inventors have realized it,
the gasket arrangement 40 may further be simply designed so as to
prevent liquid leakage during a transition between a shut state and
an open state of the liquid cooling circuit, as in embodiments
discussed later.
To start with, and as best seen in FIGS. 2, 3, the gasket
arrangement 40 may notably include a toric joint 42, which is
arranged on the cold plate 30. This toric joint 42 surrounds the
aperture of the duct 31 of the cold plate 30. This aperture is open
on that side of the cold plate 30 that is facing the wall
11.sub.-z, when the plate 30 inserted in the slit 11. In addition,
the toric joint 42 is dimensioned so as to surround the frame
outlet 14 when the assembly is in a fully open state (i.e., the
cold plate 30 is in an optimal fluid communication position, with
the duct 31 right in front of the outlet 14), to prevent liquid
leakage from the assembly 1 in that state.
The duct 31 and the frame outlet 14 shall typically have
substantially the same diameter, though small discrepancies are
possible, which can be compensated by the dimensions of the joint
42. The in-plane (inner) diameter of the joint 42 is at least equal
to the diameter of the frame outlet 14, to prevent leakage in an
open state of the circuit. Now, since the (inner) diameter of the
joint 42 may be larger than the diameter of the frame outlet 14,
one understands that there can be a finite interval of positions
(corresponding to the difference of diameters between the ring 42
and the outlet 14) of the cold plates, for which the system is
fully open.
In embodiments, the gasket arrangement 40 in fact includes two
toric joints 41, 42, FIGS. 3A, 3B. A first toric joint 41 is
located on the insert element 20, so as to surround the frame
outlet 14 when the insert element 20 is in a sealing position.
Again, the joint 41 may have a larger diameter than the outlet 14,
such that there can be a finite interval of positions of the insert
20, for which the system is fully sealed. The second joint 42 is
located on the cold plate 30, as described above. This way, the
joints 41, 42 prevent liquid leakage from the frame outlet 14 when
the system is either in a fully shut state or in a fully open
state.
Moreover, as further seen in FIGS. 3A, 3B, the gasket arrangement
40 may further comprise a simple outer joint 45-46, to prevent
leakage during transitions from one state to the other. The outer
toric joint 45-46 actually decomposes into two halves 45, 46. The
first half 45 is located on the insert element 20, on that side
facing the wall 11.sub.-z, just like the first toric joint 41. The
latter is partly surrounded by the first half 45 of the outer toric
joint 45-46 on the insert 20. The second half 46 of the outer joint
45-46 is located on the cold plate 30, on that side facing the wall
11.sub.-z, like the second joint 42. And similarly, the joint 42 is
partly surrounded by the second half 46 on that same side. Thus,
when (if) the two parts 20, 30 are in contact in the slit 11, the
two halves 45, 46 meet and form the outer toric joint 45-46 (see
FIG. 3A). In that case, the toric joints 41, 42 happen to be both
located within an inner area bounded by the outer joint 45-46,
i.e., the area subtended by the outer joint 45-46 on elements 20,
30.
With such a configuration, the toric joints 41, 42 are side-by side
on a same side of the cold plate 30 and the insert element 20
(though not necessarily in a same plane, owing to a possible shift
t between surfaces of elements 20 and 30, as discussed later), and
fully surrounded on that same side by the outer toric joint 45-46
that forms when the parts 20, 30 are in contact (FIG. 3A). The
buffer area that is defined within the outer joint 45-46 but
outside the inner joints 41, 42 define, together with the joints
41-46 (which protrude outwardly from their respective anchorage
surfaces), a buffer volume in the slit 11. Liquid can accumulate in
this buffer volume and be contained during a transition from one
state to the other, to prevent leakage during such a
transition.
The joints 41-46 are mechanical gaskets, which are typically shaped
as tori (e.g., forming a loop with a rounded cross-section). Such
joints are, for example, designed to be seated in respective
grooves formed on respective anchorage surfaces of the insert 20
and plate 30. The joints 41-46 get compressed during the insertion
of the parts 20, 30 in the slit, between their anchorage surfaces
and the opposite wall 11.sub.-z, of the slit 11, which creates a
seal at the interface. Still, the resulting friction can be
pre-determined so as to be overcome by exerting a reasonable force
on the insert and/or the plate. The joints are typically made from
elastomer materials which are able to deform (to some extent) for
the parts 20, 30 to tightly fill the slit 11 at the level of the
joints 41-46.
In terms of dimensions, the joints 41 and 42 typically have
identical or similar dimensions, with diameters that, for example,
are between 2 mm and 10 mm, and for example, of about 5 mm. Their
thickness (or height) is for, example, between 0.2 mm and 1.0 mm,
and for example, of about 0.5 mm. The joints 45 and 46 typically
have identical or similar dimensions too. Their dimensions may for
instance be chosen such that the spacing between the joints 45 and
41 and the spacing between the joints 46 and 42 is between 0.2 mm
and 2 mm, and for example, of about 0.5 mm. The apertures of ducts
14 and 31 typically have identical or similar dimensions, which are
chosen such that the diameter is slightly smaller than the diameter
of the joints 41 and 42. The diameters of such ducts, for example,
are between 2 mm and 9 mm, and for example, of about 4 mm.
In terms of materials, the joints 41, 42, 45, and 46 are typically
made from elastomer materials. Examples include: natural rubbers,
silicone rubbers, fluorosilicone rubbers, butyl rubbers,
polyurethanes, polytetrafluoroethylene (PTFE), and ethylene
propylene diene monomer (M-class) rubbers (EPDM).
As further seen in FIG. 2C, the assembly 1 may further comprise
fastening means 39, configured to help maintaining the two parts
20, 30 and, thus, the two halves 45, 46 forming the outer toric
joint 45-46, in operation (see FIGS. 1, 3A and 4). Such fastening
means 39 may for instance include hooks or other interlocking
features (e.g., a snap-fit mechanism or other integral attachment
features) that keep the two parts 20, 30 in contact after
attachment, e.g., when sliding the insert 20 and the cold plate 30
in the slit 11. In FIG. 2C, such fastening means 39 are assumed to
be realized as edges (rounded at the tips), which protrudes from
the lower side of the plate 30 and are meant to be engaged in
complementary snap-in cavities (not shown) realized in the insert
element 20.
Bulging features (not shown) may possibly be provided in the frame
10, the insert 20, or, still, the cold plate 30, in order to
further constrain the cross section of the joints 41-46, in order
to improve the sealing action. For example, bulging features may be
provided in the grooves in which the joints are received, to
further constrain the joints.
As further seen in FIGS. 2 and 4, the cold plate 30 may
advantageously include a liquid reservoir 37, as well as one or
more apertures 34. The latter are notably open in the buffer area,
i.e., on that side of the cold plate that faces the wall 11.sub.-z,
within the inner area bounded by the outer joint 45-46, but outside
the toric joints 41, 42. Such apertures 34 may easily be formed at
an edge, as depicted in FIG. 2C, so as to be surrounded by the
outer joint 45-46.
As further seen in FIGS. 2A and 2B, the liquid reservoir 37 forms
part of a third liquid coolant circuit 34-38. The reservoir 37 is
in fluid communication with the apertures 34, thanks to sections
35, 36 of the third liquid circuit portion 34-38. Note, the third
circuit 34-38 is independent from the second 31-33. Instead, the
third circuit 34-38 makes it possible to evacuate liquid coming
from the frame outlet 14 toward the reservoir 37, during a
transition from one state of the system to the other, in operation
of the assembly 1. I.e., liquid in the buffer volume defined by the
buffer area can exit through apertures 34 to reach the reservoir
37.
Note, additional ducts and/or reservoirs may possible be provided,
in the cold plate 30 or in additional parts (no shown) of the
system. In addition, adsorbing elements may possibly be provided
(not shown), e.g., within the reservoir and/or the gap defined
between the parts 20, 30 and the inner wall 11.sub.-z, to further
reduce liquid spillage. Further adsorbing elements may also be
attached to the insert element 20 or the cold plate 30. However,
such adsorbing elements would ideally have a relatively large
surface area. Thus, they may advantageously be attached to an inner
side wall of the liquid reservoir 37. Other adsorbing elements may
nevertheless be attached within the within the buffer area, i.e.,
on the insert element 20 (between the joints 45 and 41) and/or on
the cold plate 30 (between the joints 46 and 42).
As illustrated in FIGS. 2A-B, and 2D, or otherwise suggested by
FIG. 4D-E, the reservoir 37 is, for example, designed so as to
allow evaporation of liquid accumulated therein, via an edge
surface of the cold plate 30. E.g., the cold plate 30 may include a
liquid evaporation medium 38 extending along its upper edge
surface, to ease evaporation of the liquid L.sub.R (as accumulated
in the reservoir) in air. This medium may for example be a porous
medium 38 or a medium comprising capillary features (e.g.,
openings, pillars, etc.), on the upper edge surface of the cold
plate 30.
Referring back to FIG. 3B, the sealing surfaces of the insert 20
and the plate 30 may typically have to be shifted with respect to
each other in practice, e.g., due to manufacturing tolerances or
other constraints for the thicknesses of parts 20 and 30. E.g., the
face of the cold plate 30, on which the joint 42 is arranged, may
thus have to be recessed by a distance t along the normal of that
face, with respect to the face of the insert 20 on which the joint
41 is arranged. The recess gap t may thus impact the gap volume g
defined between, on the one hand, the solid frame 10, and, on the
other hand, both the insert 20 and the cold plate 30 when the
insert 20 and the cold plate 30 are both inserted in the slit 11
and in contact. There, the joints 41-46 may have different
thicknesses to compensate for this and thus all be level with each
other (on the side of the inner wall 11.sub.-z), so as to all
contact that inner wall 11.sub.-z, and ensure a homogeneous sealing
action. Note, however, that a gap volume g exists even if t=0,
owing to the residual heights of the joints 41-46. I.e., a gap
volume g is in all cases ensured by the residual heights of the
compressed joints 41-46 above their respective anchorage surfaces.
Now, if the surfaces 20, 30 are shifted with respect to each other,
then the residual heights of the joints 42, 46 need to differ from
the residual heights of the joints 41, 45 for the upper edges of
the joints to be all flush. In all cases, liquid may accumulate in
the gap volume g during transitions between the shut and open
states of the liquid cooling circuit, to prevent leakage.
As evoked earlier, each joint 41-46 may be partly embedded under
its respective anchorage surface, in order to ensure a satisfactory
anchorage of the joints. That is, all toric joints may be anchored
(e.g., thanks to grooves) on a same side of elements 20, 30.
Although same types of joints will normally be used, joints of
different section diameters may be used to compensate for a
possible thickness mismatch t between elements 20 and 30, as evoked
above. Similarly, different embedding level (resulting from
different anchorage groove dimensions or shapes) may be relied on
to compensate for the mismatch t. In all cases, the top edges of
the joints can be made level with each other, so as to ensure a
suitable fit and sealing. Still, the compressibility of the joints
may compensate for slightly non-level joints.
As further seen in FIGS. 2, 4, the cold plate 30 may further
comprise liquid cooling structures 33, which are arranged in a
volume defined at an end of the second circuit 31-33, within the
plate 30. Such structures 33 may for example be realized as
structures protruding from a basis surface (defined within the
plate 30), so as to effectively increase the surface area
contacting the liquid accumulated in the plate and improve the
cooling.
In embodiments shown in FIGS. 1-4, only one lateral side of the
parts 10-30 is shown and, as said earlier, the assembly 1 need not
be symmetric. However, it for example is, be it to ease insertion
of the parts 20, 30 within opposite slits 11, 11a, as discussed now
in reference to FIG. 5.
As seen in FIG. 5, each of the solid frame elements 10, 10a and the
cold plate 30 may be symmetrically designed (with respect to the
plane (x, z). Note, the solid frame may further include a chassis
(not shown, below the sabots 10, 10a), so as to substantially have
a U-section in the plane (x, y). In such a symmetric design, at
least one pair of opposite slits 11, 11a are meant to receive the
parts 20, 30. Also, in that case, the first liquid coolant circuit
12-14, 12a-14a now comprises two subsections, i.e., the subsections
12-14 and 12a-14a subtended by the two frame outlets 14, 14a, which
are oppositely located on lateral edges of the frame 10.
Note, a single insert element 20 could be used to shut the liquid
circuit, by inserting this element in a pair of opposite slits.
However, in order to maximize the useful area of the plate 30, best
is to use a suitably profiled insert or, even, two insert elements
20 (as in FIG. 5), where each of the elements 20, 20a can be
inserted in a respective one of the opposite slits 11, 11a, on each
lateral side of the frame elements 10, 10a. Both inserts 20, 20a
may reach their respective sealing positions, to seal the circuit
portions 12-14 and 12a-14a, which are thus disconnected. I.e., this
result in a shut state of the corresponding section of the liquid
cooling circuit, where the first and second insert elements 20, 20a
seal a respective frame outlet 14, 14a.
As further suggested by the symmetric arrangement of FIG. 5, the
second liquid coolant circuit now comprises two ducts 31, 31a,
which are both open on a same side of the cold plate 30. The cold
plate 30 can be pushed further down in the slits 11, 11a, so as to
push each insert 20, 20a, for the latter to leave their sealing
positions. This way, the cold plate 30 can reach a fluid
communication position, which defines an open state, in which each
duct 31, 31a is (at least partly) vis-a-vis a respective frame
outlet 14, 14a, to connect subsections 12-14, 12a-14a of the first
circuit via the second circuit 31-31a.
Using symmetric designs such as depicted in FIG. 5 allows a simple
operation of the assembly 1. I.e., the circuit sections can be
connected/disconnected by gently pushing/pulling the plate 30. IC
packages (not shown) mounted on the respective plates 30 can thus
easily be hot plugged/unplugged, even if the active cooling circuit
is switched on. Also, in a symmetric design, two or more reservoirs
37, 37a may be provided in the cold plate 30, to buffer liquid on
each side and prevent leakages during the mounting or de-mounting
of the cold plate 30. One or more liquid cooling structures 33 may
be present too, which have the same function as described
earlier.
Furthermore, the first circuit subsections 12-14, 12a-14a and the
second circuit portion 31-33 may, once connected, form a loop,
corresponding to a subsection of a larger liquid cooling circuit,
it being noted that the same lateral channels 12, 12a may service
several loops, each corresponding to a respective pair of opposite
slits 11, 11a. The overall circuit may be a closed cooling circuit,
wherein liquid is recirculated to cool the plates 30. In variants,
the overall circuit may be open, whereby liquid would be filled at
one of the channels 12 and evacuated from the other, opposite
channel 12a.
Although the discussion so far was merely circumscribed to the
description of the operation pertaining to a single plate 30,
embodiments of the present invention will likely involve several
pairs of opposite slits 11, 11a, allowing connection/disconnection
of several cold plates 30, by means of several inserts 20 (or pairs
of inserts 20, 20a). This way, dense packaged electronics may be
cooled and yet easily connected/disconnected, following the same
principles as discussed above.
In that respect, and according to another aspect, the invention may
be embodied as a hardware system (e.g., a computerized system),
where the system comprises a disconnect assembly 1 such as
described herein, as well as packaged electronics 50 mounted on the
cold plate 30, or somehow attached to the plates 30, to thereby
cool down the packaged electronics 50, in operation of the system.
The packaged electronics may for instance include, each, one or
more chips, e.g., memory chips, mounted on a printed circuit board
(PCB) to form an integrated-circuit (IC) package 50. Sockets (not
shown), solder pads (not shown) or other interconnects, will ensure
proper connections of the IC packages. Such interconnects may,
however, be provided on other components of the system, outside the
assembly 1.
Referring to FIGS. 4A-E, a method of operation of an assembly 1 or
a system such as described above is now briefly discussed, which
concerns another aspect of the invention. Main aspects of this
method have already been evoked earlier in the description of the
components 10-30. Essentially, this method requires to first insert
an insert element 20 in a slit 11, so as for the insert element 20
to reach a sealing position, in which the element 20 seals a frame
outlet 14, see FIG. 4A, or a pair of outlets 14, 14a, in a
symmetric design. Thus, liquid L.sub.1 coming from the lateral
channel 12 remains confined in the conduit 13. I.e., this position
defines a shut state of the circuit section formed by the first and
second liquid cooling subsections. Then, a cold plate 30 is
inserted in that same slit 11, see FIG. 4A-B.
When the plate 30 comes in contact with the insert 20 (FIG. 4B), it
starts pushing the insert element 20, such that the latter
progressively leaves its sealing position, FIG. 4C. There, liquid
L.sub.1 starts filling the buffer volume between the inner 41, 42
and outer joints 45, 46. Thanks to apertures 34 (see FIG. 2),
liquid L.sub.R starts filling the third circuit section 34-38 and
thus reaches the reservoir 37 (FIG. 4C).
Pushing the cold plate 30 and insert 20 further down, the cold
plate 30 reaches a position that already enables fluid
communication, FIG. 4D. That is, the duct 31 and outlet 14 are not
perfectly in via-a-vis yet but, still, liquid L.sub.2 can already
pass into the second circuit portion 31-33. This new position
defines a partly open state of the cooling circuit and liquid
L.sub.2 can notably reach the cooling structures 33 provided in the
plate 30. Meanwhile, liquid L.sub.R has filled the reservoir 37 and
starts evaporating, thanks to an evaporation medium 38.
Pushing further down, the cold plate 30 reaches a plain fluid
communication position, FIG. 4E, which defines a fully open state
of the cooling circuit (the duct 31 is now perfectly vis-a-vis the
frame outlet 14). Liquid L.sub.2 is still filling the cooling
compartment 33, while residual liquid L.sub.R in the reservoir 37
is being sucked into the medium 38 (e.g., by capillarity), which
liquid can thus evaporate through the medium 38, until no liquid
remains. No additional liquid is being drained through the third
circuit 35-37 at this point as the duct 31 faces the frame outlet
14, which is sealed by the joint 42. As said earlier, liquid
L.sub.2 that fills the cooling compartment 33 may further reach a
complementary circuit portion 12a (on the opposite side of the
frame 10) and possibly be recirculated, if the overall circuit
forms a closed-loop.
A disconnect assembly for active cooling of packaged electronics,
in embodiments, includes a solid frame that comprises a slit and a
first liquid coolant circuit leading to a frame outlet defined in
an inner wall of the slit. The assembly further includes an insert
element, which is insertable in the slit so as to reach a sealing
position. The latter defines a shut state, in which the insert
element seals the frame outlet. The assembly includes a cold plate,
which comprises a second liquid coolant circuit with a duct open on
a side of the cold plate. The cold plate can be inserted in the
slit, so as to push the insert element, for the latter to leave the
sealing position and the cold plate to reach a fluid communication
position. This position defines an open state, in which the duct is
vis-a-vis the frame outlet, to enable fluid communication between
the first liquid coolant circuit and the second liquid coolant
circuit. Related devices, systems and methods of operation may be
provided.
While the present invention has been described with reference to a
limited number of embodiments, variants and the accompanying
drawings, it will be understood by those skilled in the art that
various changes may be made and equivalents may be substituted
without departing from the scope of the present invention. In
particular, a feature (device-like or method-like) recited in a
given embodiment, variant or shown in a drawing may be combined
with or replace another feature in another embodiment, variant or
drawing, without departing from the scope of the present invention.
Various combinations of the features described in respect of any of
the above embodiments or variants may accordingly be contemplated,
that remain within the scope of the appended claims. In addition,
many minor modifications may be made to adapt a particular
situation or material to the teachings of the present invention
without departing from its scope. Therefore, it is intended that
the present invention not be limited to the particular embodiments
disclosed, but that the present invention will include all
embodiments falling within the scope of the appended claims. In
addition, many other variants than explicitly touched above can be
contemplated.
* * * * *
References